Scribd
Upload
189 views

Analysis Services DevOps Using Tabular Editor

This document discusses using Tabular Editor, Git, and Azure DevOps for Analysis Services DevOps. It describes setting up a Tabular model in source control using Git, creating build pipelines in Azure DevOps for continuous integration, and implementing branching strategies and workflows for features and environments. The document also covers dealing with conflicts, advanced patterns like the master model pattern, and automated deployment of models through release pipelines in Azure DevOps.

Uploaded by

arjun.ec633
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
189 views

Analysis Services DevOps Using Tabular Editor

This document discusses using Tabular Editor, Git, and Azure DevOps for Analysis Services DevOps. It describes setting up a Tabular model in source control using Git, creating build pipelines in Azure DevOps for continuous integration, and implementing branching strategies and workflows for features and environments. The document also covers dealing with conflicts, advanced patterns like the master model pattern, and automated deployment of models through release pipelines in Azure DevOps.

Uploaded by

arjun.ec633
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 23

Analysis Services DevOps using

Tabular Editor,
git and Azure DevOps

Daniel Otykier / Kapacity


Session scope
Tabular models and Source Control (git)
Topic branches and workflows
Dealing with conflicts
Automated builds (Azure DevOps)
Automated unit tests (Azure DevOps)*
Automated deployment (Azure DevOps)
Continuous Integration (Combining all of the above)

*Not covered in this talk


Motivation (why DevOps?)

Work simultaneously on multiple features


Work in teams of developers
Support multiple environments
Avoid merge conflicts / manual merging headaches
Enforce code review processes (“pull requests”)
Automate testing and avoid regressions
Easier and more consistent deployments
Demo
Getting started
Prerequisites

Tabular Editor
setx PATH "%PATH%;C:\Program Files (x86)\Tabular Editor\"
Instance of Analysis Services
Azure DevOps account
Local installation of Git (included with Visual Studio)
“Save to Folder” + Git =

Save to Folder
Split model into smaller files
Remove timestamps and inferred metadata
No metadata ordering conflicts
Git
Uses directory snapshots
no need to manually “include” items like in TFVC
Prefers many small files over few big files
Combine features and resolve conflicts using standard diff tools,
instead of specialized tools for Tabular Models (BISM Normalizer)
Step 1) Prepare your Tabular Model metadata

Load your Model.bim file in


Tabular Editor
(alternatively) Load your model
from an existing database
Go to File > Preferences >
“Current Model”
Ensure settings are as seen on
the right
Save the model using File >
Save as Folder...
Step 2) Set up your Git repository
One or multiple Git repositories? (Discussion)
For demo purposes:
Use default repository in new Azure DevOps project...
...or create a new repository in existing project
Add a “VisualStudio” .Gitignore file

Many ways to add your Tabular Model metadata to Git:


“Upload files” within the Web UI, then clone to a new local
repository (note: doesn’t support Folder upload)
Clone the repository, add the files locally, commit and push
Init local repository, add remote, add files, commit and push
Use Git commandline or Visual Studio (“Open folder”) if
preferred
Step 3) Determine branching strategy
https://git-scm.com/book/en/v1/Git-Branching-Branching-
Workflows MASTER
Deployment 1 Deployment 2

One possible approach:


Scrapping
feature 1

NEXT
MASTER = Production DB Hotfix

NEXT = Test DB
TOPIC1
TOPIC1 = New feature 1
TOPIC2 = New feature 2 TOPIC2

...
TOPIC3
Each topic = Separate workspace DB
Step 4) Create build pipelines

Download and configure build agent if requiring access to on-prem


ressources
(Otherwise, may use Hosted Agent pools)
Build agent SA needs admin access to SSAS instance (unless you specify
credentials when connecting)
Build pipelines are extremely flexible! They can do ANYTHING!
...but with great power, comes great responsibility! Just because you could
doesn’t mean you should...
Daily workflow
Typical daily workflow (after feature branch creation)

Merge from TEST to your Feature branch and deploy to DEV DB


Open your DEV DB in Tabular Editor and work work work...
Hit CTRL+S to save to the DEV DB, and test your changes in Excel / Power
BI / SSMS
Periodically use “Save to Folder...” to save your work into the current
working directory
Consider committing to Git as well (frequent commits are a GOOD thing!)
Might as well push to Azure DevOps, so that you don’t lose any work if your laptop
is stolen/goes into perpetual BSOD/takes a brick to the face/etc... 
When the feature is done, pull request into the TEST branch
Dealing with conflicts

May occur when merging from TEST into your current Feature
branch
I.e. someone else changed something related to what you’re currently
working on
Easiest to do within Visual Studio
Advanced patterns
Master Model Pattern

Useful when maintaining 2+ models with a reasonable amount of


functional overlap, such as:
Shared dimensions
“Process once”-scenarios
Repeated complexity
Helps consolidate your code base
https://Github.com/otykier/TabularEditor/wiki/Master-model-
pattern
Automated deployment

Create “Release definition” based on a Build definition


Define stages, for example: Test, Pre-prod, Prod
For each stage, define deployment tasks (similar to build pipeline)
Use variables!
Can be set to “sensitive” for connection strings, passwords, etc.
Continuous integration

Set up triggers on build- and release pipelines for end-to-end


automation
For example:
After every commit, run BPA, SchemaCheck and deploy to build server
Run unit tests
If all succeeds, deploy to test server and inform stakeholder
If stakeholder approves, deploy to prod
Conclusion

You need Tabular Editor!


You need Git
You need Azure DevOps
Command-line scripts are oldschool cool
Next steps:
Unit testing using PowerShell or NBI
CI/CD for your Tabular data source
Blog series coming soon:
https://tabulareditor.github.io/2019/02/20/DevOps1.html
Thank you!

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy